Hydrophobic Polyoxometalate-Based Metal-Organic Framework forEfficient CO2 PhotoconversionXiao-Xin Li,†,‡ Jiang Liu,‡ Lei Zhang,‡ Long-Zhang Dong,‡ Zhi-Feng Xin,‡,§ Shun-Li Li,‡

Xue-Qing Huang-Fu,‡ Kai Huang,*,† and Ya-Qian Lan*,‡

†School of Chemistry and Chemical Engineering, Southeast University, Nanjing 211189, P. R. China‡Jiangsu Collaborative Innovation Centre of Biomedical Functional Materials, Jiangsu Key Laboratory of New Power Batteries,College of Chemistry and Materials Science, Nanjing Normal University, Nanjing 210023, P. R. China§Institute of Molecular Engineering and Applied Chemistry, Anhui University of Technology, Ma’anshan 243002, P. R. China

*S Supporting Information

ABSTRACT: A novel polyoxometalate (POM)-based metal-organic framework,TBA5[P2Mo16

VMo8VIO71(OH)9Zn8(L)4] (NNU-29), was in situ synthesized and applied

into CO2 photoreduction. The selection of porous material containing a reductive POMcluster is considered to be helpful for CO2 reduction; meanwhile, a hydrophobic-group-modified organic ligand enables NNU-29 to exhibit good chemical stability and restrainshydrogen generation to some extent. In the photocatalytic CO2 reduction, the yield ofHCOO− reached 35.2 μmol in the aqueous solution with selectivity of 97.9% after 16 h.

KEYWORDS: polyoxometalate-based metal-organic frameworks, hydrophobicity, structural stability, heterogeneous catalyst,carbon dioxide photoreduction

1. INTRODUCTION

The burning of fossil fuels leads to excessive carbon dioxideemission into the atmosphere, causing serious environmentaland energy problems that need to be solved urgently.1−3

Photocatalytic reduction of CO2 into value-added chemicals orhydrocarbon fuels in water to simulate natural photosynthesiscan slow down the process of energy consumption and reduceCO2 concentration.

4−8 However, the chemical inertness of theCO2 molecule makes it difficult to be activated and converted,which requires effective catalysts with reducibility to participatein this process and accomplish it. Moreover, as a competitivereaction, the hydrogen evolution reaction always reduces theselectivity of the photocatalytic reaction. Not only does it wearout portion of the photogenerated electrons for CO2conversion, but also it is disadvantageous for further separationand purification of the target product. Therefore, it is a goal toconstruct reductive photocatalysts to convert CO2 with highselectivity.Metal-organic frameworks (MOFs) have been widely

investigated in the fields of gas adsorption and catalysis overthe decades due to their structural controllability and opencatalytic sites.9−12 Meanwhile, the appropriate spatial structureof MOFs can help understand the catalysis mechanism at themolecular level.13−16 It is noticed that the secondary buildingunits, which always play the role of a catalytic center, directlydetermine the efficiency of the catalytic reaction, mostly.Introducing reductive nodes to construct MOFs can result in a

great performance in some reduction reactions.17−20 It is well-known that polyoxometalates (POMs) and their derivativesshow outstanding properties in plenty of catalytic reactions, inwhich POMs function as high-efficiency catalytic sites withsplendid redox ability and semiconductor characters.21−24

Nevertheless, simple POM clusters are often soluble in water,so it is difficult to recycle them and they may causeenvironmental pollution. Therefore, stability is the prerequisitefor POM-based derivatives as heterogeneous catalysts. It isnoted that when connecting POMs with linkers to constructpolyoxometalate-based metal-organic frameworks (POMOFs),they can not only retain the advantages of POMs but also keepstable in aqueous solution.25 This indicates a potential methodto select reductive POMs like {Zn4PMo8

VMo4VI} or

{Mo6VO12(OH)3(HPO4)3(PO4)}

26 to synthesize POMOFsas photocatalysts, which may have beneficial effects on CO2reduction.Based on the above considerations, we combined POM and

a hydrophobic ligand to synthesize a stable POM-based metal-organic framework, TBA5[P2Mo16

VMo8VIO71(OH)9Zn8(L)4]

(NNU-29; L2− is 4,4′-((((perfluoropropane-2,2-diyl)bis(4,1-phenylene))bis(oxy))bis(methylene))dibenzoate anion andTBA+ is the tetrabutylammonium ion). With the help of a

Received: March 4, 2019Accepted: June 26, 2019Published: June 26, 2019

Research Article

www.acsami.orgCite This: ACS Appl. Mater. Interfaces 2019, 11, 25790−25795

© 2019 American Chemical Society 25790 DOI: 10.1021/acsami.9b03861ACS Appl. Mater. Interfaces 2019, 11, 25790−25795

Dow

nloa

ded

via

NA

NJI

NG

NO

RM

AL

UN

IV o

n Se

ptem

ber

1, 2

019

at 0

9:08

:27

(UT

C).

See

http

s://p

ubs.

acs.

org/

shar

ingg

uide

lines

for

opt

ions

on

how

to le

gitim

atel

y sh

are

publ

ishe

d ar

ticle

s.

fluorine-containing ligand, NNU-29 shows strong hydro-phobicity and chemical stability. It is noticed that introducingthe {Zn4PMo8

VMo4VI} unit with strong reducibility into the

structure enables NNU-29 to perform efficient heterogeneousphotocatalytic conversion of CO2 to HCOOH in water. TheHCOOH production is 220 μmol g−1 h−1 along with theselectivity of nearly 98%. The existence of a fluorine-containingligand avoids the attack from water to generate H2 to somedegree.

2. EXPERIMENTAL SECTION2.1. Synthesis of TBA5[P2Mo16

VMo8VIO71(OH)9Zn8(L)4] (NNU-

29). A mixture of Na2MoO4·2H2O (0.600 g, 2.50 mmol), H3PO3(0.020 g, 0.25 mmol), ZnCl2 (0.20 g, 1.47 mmol), and H2O (6.0 mL)was stirred for 1 h, resulting in a milky white suspension. Then,molybdenum powder (60 mg, 0.62 mmol) and H2L (0.063 g, 0.10mmol) were introduced. TBAOH (25 wt %) in methanol (0.1 mL,0.09 mmol) and HCl (0.2 mL, 6.0 M) was added to adjust the pH ofthe aqueous suspension. The solution was mixed under vigorousstirring, ultrasound-treated for approximately 30 min, and transferredand sealed in a 15 mL Teflon-lined stainless steel containermaintained at 180 °C for 72 h and then cooled to room temperaturefor more than 24 h. Dark red rhombus crystals of NNU-29 (FigureS2) were separated and washed with Millipore water, which werecollected in 40% yield based on H2L. The CCDC reference number is1893141. Synthesis procedures of H2L and Zn4-ε-Keggin are given inthe Supporting Information.2.2. Photocatalytic Test. The photocatalytic CO2 reduction

experiments were carried out on an evaluation system (CEL-SPH2N,CEAULIGHT, China) in a 100 mL quartz container. A 300 W xenonarc lamp with a UV-cutoff filter (420−800 nm) was utilized as theirradiation source. The photocatalysts (NNU-29, Zn4-ε-Keggin orH2L) were dispersed in 50 mL solution, and pre-degassed with CO2(99.999%) for 30 min to remove air before irradiation. The reactionmixture was kept stirred constantly with a magnetic bar to ensurecomplete mixing of the photocatalyst particles in suspension. Thetemperature of the reaction was maintained at 20 °C by a circulatingcooling water system.

3. RESULTS AND DISCUSSION

X-ray crystallography reveals that NNU-29 crystallized in themonoclinic space group C2/c and with two formula units (Z =2) per unit cell (Table S2). NNU-29 consists of the ε-{Zn4PMo12O40} cluster as the inorganic building node, thehydrophobic fluorine-based ligand as the linker, and TBA+

ions. The secondary building unit can be considered as ε-{PMo12O40} capped by four Zn ions in tetrahedralsymmetry.27−30 There are four Mo(VI) and eight Mo(V) inone {Zn4PMo12O40} cluster, indicating that the Keggin node isa reductive POM (Figure S14b). Each grafted Zn ioncoordinates with the carboxylic oxygen from a flexible L2−

ligand to generate a two-dimensional (2D) network, which isshown in Figure 1a. Besides, regarding the ε-{Zn4PMo12O40}cluster as a 4-connected node, the skeletons of NNU-29feature a topology with the Schlafli symbol of 44 · 62, whichreveals the topology of the sql code. Figure 1b shows the layernetwork along the c-axis. In fact, the authentic layer containsthree independent interpenetrating frameworks, which areshown in Figure S4.The contact angle of H2L is measured to be 133.68° (Figure

S1), which shows that it is a hydrophobic ligand. What isinteresting and worth noting in the structure is that all of thelocations of trifluoromethyl are around the POM node, whichcan protect the hydrophilic POMs, avoiding the attack fromH2O. Along the b-axis, POMs and trifluoromethyl are likely tobe two kinds of layers interlaced together to form a sandwich-like structure (Figure 1c). In the view along the c-axis, fourrows of POMs leave a rhombic channel, which is full oftrifluoromethyl (Figure 1d). In the interspace, TBA+ ions act ascountercations to balance the charges. Because of the above-mentioned factors, NNU-29 shows hydrophobic character-istics. The result of contact angle measurement was 122°,which evidences NNU-29 as hydrophobic (Figure 2a).The introduction of hydrophobic ligands results in NNU-29

exhibiting good hydrophobic properties and acid−base

Figure 1. Summary of the structure of NNU-29: (a) Coordination environments of NNU-29; all hydrogen atoms are omitted for clarity. P, yellow;Mo, light blue; Zn, violet; C, gray; O, red; N, blue; and F, green. (b) The single 2D layer network along the c-axis. (c) POM layers andtrifluoromethyl layers form a sandwich-like structure along the b-axis. (d) A rhombic channel full of trifluoromethyl along the c-axis.

ACS Applied Materials & Interfaces Research Article

DOI: 10.1021/acsami.9b03861ACS Appl. Mater. Interfaces 2019, 11, 25790−25795

25791

stability. Phase purity of single-crystal NNU-29 wasdetermined by well-matched powder X-ray diffraction(PXRD) patterns (Figure 2b). The chemical stability measure-ments were performed by soaking prepared samples in aqueoussolution for 48 h under different conditions. The consistentpeaks reveal that NNU-29 can remain crystalline in 4 M HCland 0.01 M NaOH solution. The structural integrity of NNU-

29 in triethanolamine (TEOA) solution (1:4, v/v) provideslong-term durability during CO2 photoreduction process ascatalysts directly. This rigid structure exhibits superiorchemical stability, which is fairly rare among most MOFsand POMOFs.The thermal stability of NNU-29 was examined through

thermogravimetric analysis. As shown in Figure S7, there is nosignificant change in thermal weight before 300 °C, indicatingthat NNU-29 has good thermodynamic stability. The 46.71%weight loss of NNU-29 from 300 to 680 °C is attributed to thesplitting of TBA+ ions and the release of L2− ligand, calculatedto be 46.44%.TBA4[PMo8

VMo4VIO37(OH)3Zn4]Cl4 (Zn4-ε-Keggin), the

inorganic node of NNU-29, was further synthesized,31 and itsproperties were studied. Suggested by the ultraviolet−visible(UV−vis) results shown in Figure 2c, NNU-29 reveals broadUV−vis light absorption covering the entire UV−visible lightregion. On comparing the absorption of Zn4-ε-keggin andpure H2L, it can be firmly believed that the absorption ismainly contributed by the ε-{Zn4PMo12O40} cluster. The bandgaps (Eg) of NNU-29 and Zn4-ε-keggin were furthercalculated from original UV−vis data with the Kubelka−Munk formula: (αhv = C(hv − Eg)

2). In Figure 2d, theintersection of the tangent lines from the curve and horizontallines shows that the band gaps are 1.67 and 1.91 eV for NNU-29 and Zn4-ε-keggin, respectively. To confirm the position ofthe valence band (VB) of NNU-29, ultraviolet photoelectronspectroscopy (UPS) was performed (Figure 2e). The resultwas ultimately calculated to be 5.73 eV by deducting the widthof the He I UPS spectra from the excitation energy (21.22 eV).Then, the position of the conduction band was evaluated to be4.06 eV by Ev − Eg. What is revealed in Figure 2f is the bandstructure of NNU-29. The values of bending coefficient (CB)and VB were converted to the normal hydrogen electrode

Figure 2. (a) Contact angle measurement of NNU-29. Both the leftand right angles are 122°. (b) Powder X-ray diffraction (PXRD)patterns of NNU-29 in different solutions compared with simulatedcurves. (c) UV−vis absorption of NNU-29 (black curve), Zn4-ε-Keggin (red curve), and H2L (blue curve). (d) (αhν)2 vs hν curve ofNNU-29 (black curve) and Zn4-ε-Keggin (red curve). (e) Ultravioletphotoelectron spectroscopy (UPS) spectra of NNU-29. (f) Bandstructure diagram for NNU-29.

Figure 3. (a) Amount of HCOO− produced as a function of the time of visible-light irradiation over NNU-29, Zn4-ε-keggin, and H2L, Thereaction with each photocatalyst (10 mg) in the H2O/TEOA (4:1 v/v, 50 mL) solution was irradiated using a Xe lamp filtered to produce light inthe range of 420−800 nm; (b) the evolution of HCOO− produced in five consecutive runs (recovery of the used catalyst and then redispersed in afresh photosensitizer solution for 16 h in each run) of the photocatalytic CO2 reaction over NNU-29; (c) the effect of different quantities of NNU-29 on the yield (left) and rate (right) of HCOO− from the CO2 photoreduction system; and (d) 13C NMR spectra of the product obtained fromthe reaction with 13CO2 (red) and

12CO2 (black).

ACS Applied Materials & Interfaces Research Article

DOI: 10.1021/acsami.9b03861ACS Appl. Mater. Interfaces 2019, 11, 25790−25795

25792

(NHE) in volts (right y-axis). The reduction potential of CO2/HCOOH is lower than the conduction band of NNU-29,which indicates that it can be a photocatalyst for CO2-to-HCOOH reduction theoretically. To further confirm thesemiconductor band structure, the Mott−Schottky plotmeasurement on NNU-29 was implemented at frequenciesof 1000, 1500, and 2000 Hz. As shown in Figure S8, theobtained C−2 curves with the x-axis were intersected at −0.76V vs Ag/AgCl. Then, the position of the CB was furthercalculated to be −0.66 V vs NHE, which agreed with the resultof UPS.In addition, NNU-29 has a distinguished transient photo-

current response under visible-light irradiation when the lightis “turn on” or “turn off” (Figure S9). The sensitivephotocurrent response of NNU-29 represented a highseparating efficiency of photoinduced electron−hole pairs. Itis important to note that NNU-29 has been proven to be asemiconductor material with a conduction band-edge potentialthat matched the theoretical reduction potential required byHCOO−.In the system of CO2 photoreduction, triethanolamine

(TEOA) as the sacrificial agent and [Ru(bpy)3]Cl2·6H2O(11.2 mg, 0.0134 mmol) as the supporting photosensitizer(PS) were introduced. It is a widely approved approach toimprove the performance.32,33 Because of the appropriatelowest unoccupied molecular orbital (LUMO) positionsbetween catalysts and PS (Figure S10), photo-motivatedelectrons were allowed to transfer from PS to catalysts. As weexpected, NNU-29 reveals noteworthy photocatalytic perform-ances for CO2 reduction. By testing the yield of HCOO− at 4,8, 12, and 16 h, respectively, we achieve a roughly linearrelationship between the target product and the reaction time(Figure 3a). The amount grew to 35.20 μmol in 16 h, whichcorresponded to a turnover number (TON) (Table 1) of 28.09

for HCOO− formation. There were no any other productsdetected in the liquid. A little amount of CO (0.18 μmol) andCH4 (0.58 μmol) gas production (Figure S11) was detected bygas chromatography (Figure S17). In addition, no affiliated H2produced during the entire reaction (Figure S18). The resultsabove indicated a high selectivity up to 97.9% for CO2

photoreduction. In addition, NNU-29 can be easily recoveredfrom the test reaction solution by centrifugation. Afterremoving the supernatant liquid, the catalyst was washedwith water three times. The sample was dried in the air for thenext round. As shown in Figure 3b, NNU-29 can keep theoriginal activity of ca. 90% after five cycles. The PXRD patterns(Figure S12), Fourier transform infrared spectroscopy (FigureS13), and X-ray photoelectron spectroscopy (Figure S14a)showed no significant changes before and after the cycleexperiment, which proved the structural stability of NNU-29again.The effect of the quantity of NNU-29 on the photochemical

performance of the system was further studied. As shown inFigure 3c, the amount of HCOO− produced was increasingalong with the increase in the amount of NNU-29. When 5 mgof NNU-29 was introduced into the system, 15.90 μmol ofHCOO− was produced corresponding to a rate of 198.75 μmolg−1 h−1. When the amount increased to 10 mg, the rate ofHCOO− formation reached the highest value of 220 μmol g−1

h−1. Even though the yield of HCOO− enhanced along withthe greater quality of photocatalysts, the formation rate ofHCOO- decreased sharply. It could be ascribed to the rate-determining step of the process, which was the transfer ofphotoelectron from PS to NNU-29. The redundant catalystdid not serve completely due to the limited generation ofphotoelectron.34

To investigate the active roles of CO2 photoreduction, aseries of contrast experiments were conducted, and the resultsare listed in Table 1. When the experiment was carried out inthe dark (entry 2), no any products were detected either in gasor liquid, indicating that it is a light-driven catalytic reaction.When the experiment was carried out in lack of NNU-29, noHCOO− but little CO and H2 could be observed (entry 3). Inthe absence of [Ru(bpy)3]Cl2·6H2O, the efficiency of CO2reduction decreased dramatically and only trace of CO can bedetected (entry 4). The result showed that the photoelectronsexcited from PS and NNU-29 might be the active centers inthe photocatalysis process. To further confirm where thecatalysis reaction happens, the properties of pure Zn4-ε-keggin(entry 5) or H2L (entry 6) were examined. Contrastively, Zn4-ε-keggin revealed performance of 7.5 μmol production,indicating that the exact catalysis center is the ε-{Zn4PMo12O40} cluster. In the homogeneous system consist-ing of H2L, [Ru(bpy)3]Cl2·6H2O, and TEOA, almost the sameyield as that in entry 4 showed the insertion of ligand. Then,the physical mixture of Zn4-ε-keggin and H2L was used toreplace NNU-29 (entry 7). The result showed an obvious butweaker activity than that of NNU-29, revealing that thefluorine-containing network of NNU-29 promoted the effecton CO2 conversion.35,36 The participation of CO2 in thereaction was also studied by replacing CO2 with Ar (entry 8).There is no carbonaceous substance detected, which indicatesthat the HCOO− was transformed from CO2. Furthermore, anisotopic 13CO2 experiment was carried out. The production ofHCOO− was analyzed by 13C NMR spectroscopy. As shown inFigure 3d, when using 12CO2 as a carbon source, the peakscorresponding to TEOA and deuterium-dimethyl sulfoxidewere easily found in the 13C NMR spectrum from the filtrateafter reaction. In the range from 158 to 166 ppm, two weakpeaks at 158.2 and 159.9 could be due to HCO3

− and CO32−,

respectively, for the large solubility of CO2 and the alkalineenvironment due to the existence of TEOA to form these ions.When 13CO2 was employed in the reaction system, a new and

Table 1. Research of Reaction Conditions for NNU-29a

entryHCOOH(μmol)

CO(μmol)

CH4(μmol)

H2(μmol) TON

selectivity(%)

1 35.2 0.18 0.58 n.d. 28.09 97.92bD n.d. n.d. n.d. n.d.3c n.d. 0.76 n.d. 0.684d n.d. 1.85 n.d. n.d.5e 7.5 2.45 0.25 0.4 3.39 70.76f 0.8 n.d. n.d. n.d. 0.0487g 6.4 1.56 0.19 n.d. 78.58h n.d. n.d. n.d. n.d.

aReaction conditions: NNU-29 (10 mg, 1.28 μmol), [Ru(bpy)3]Cl2·6H2O (11.2 mg, 13.4 μmol), solvent (50 mL, H2O/TEOA, 4/1), CO2(1 atm), λ ≥ 420 nm, 20 °C, 16 h reaction time; n.d. = not detectable;turnover number (TON) = (n(HCOOH + CO + CH4 + H2))/n(photocatalyst); selectivity = (n(HCOOH))/(n(HCOOH + CO+CH4 + H2)) × 100%, where n(HCOO−) and n(photocatalyst) werethe amounts of HCOO− (mol) and the catalyst (mol), respectively.bDark condition. cNo catalyst. dNo [Ru(bpy)3]Cl2·6H2O.

eZn-ε-Keggin replaced NNU-29. fH2L replaced NNU-29. gPhysical mixtureof Zn4-ε-keggin and H2L to replace NNU-29. hAr replaced CO2.

ACS Applied Materials & Interfaces Research Article

DOI: 10.1021/acsami.9b03861ACS Appl. Mater. Interfaces 2019, 11, 25790−25795

25793

obvious peak appeared at 163.9 ppm, which corresponds to theHCOO− ion, which strongly indicates that the producedHCOO− indeed transformed from CO2. Overall, these resultshave demonstrated the main role of NNU-29 acting as a CO2redox promoter in the photocatalytic reduction system. Forcomparison, we have listed the performance of a series of othermetal-organic compounds as the photocatalysts for CO2-to-HCOOH conversion in the Supporting Information (TableS6).Upon analyzing the experimental facts and theoretical

calculation (Figure S21), a speculative reaction mechanismin regard to photocatalytic CO2-to-HCOOH reduction overNNU-29 was proposed (Figure 4). First, the photosensitizer

[Ru(bpy)3]2+ absorbs visible light. Photoelectrons are excited

from highest occupied molecular orbital to LUMO and thentransferred to the catalyst NNU-29 through the matchedLUMO positions. NNU-29 is like an electronic sponge andreceives photoelectrons due to the valence change of multipleMo ions. Second, reductive POM units play a vital role incapturing and reducing CO2 and then convert back intooriginal photocatalysis. TEOA as an electron donor consumesthe photoinduced holes produced in the valence band of[Ru(bpy)3]

2+.

4. CONCLUSIONSIn summary, a novel POMOF (NNU-29) with a two-dimensional sandwich plane structure was reported. Due tothe introduction of a hydrophobic ligand, NNU-29 exhibitshigh chemical stability. ε-{Zn4PMo12O40}, a cluster withoutstanding reducibility, prompts NNU-29 to show out-standing property as a heterogeneous catalyst for CO2photoreduction under visible-light irradiation. Meanwhile,hydrophobicity depresses the evolution of hydrogen to somedegree, which guarantees the high selectivity of CO2-to-HCOO− conversion (97.9%). The relation between thestructure and properties of NNU-29 is a successful case fordesigning efficient catalysts with contemporary selectivity. It isexpected that a rational tactics to design a stable structure withreductive inorganic clusters and functional ligands byreasonable synthesis strategies will have a significant impacton increasing the activity and selectivity in CO2 photo-reduction.

■ ASSOCIATED CONTENT*S Supporting InformationThe Supporting Information is available free of charge on theACS Publications website at DOI: 10.1021/acsami.9b03861.

Crystallographic data, NNU-29 (CIF)

Materials and measurements; analytical techniques;single-crystal X-ray diffraction; electrochemical measure-ments and computation; synthesis of L(OMe)2, H2L;the Mott−Schottky plot and the photocurrent of NNU-29; PXRD before and after reactions of NNU-29 andZn4-ε-keggin;

13C NMR spectrum; ion chromatographand gas chromatography analysis after CO2 photo-reduction; and theoretical calculation detail (PDF)

■ AUTHOR INFORMATIONCorresponding Authors*E-mail: huangk@seu.edu.cn (K.H.).*E-mail: yqlan@njnu.edu.cn (Y.-Q.L.).ORCIDLong-Zhang Dong: 0000-0002-9276-5101Kai Huang: 0000-0002-5768-4189Ya-Qian Lan: 0000-0002-2140-7980Author ContributionsThe manuscript was written through contributions of allauthors. All authors have given approval to the final version ofthe manuscript.NotesThe authors declare no competing financial interest.

■ ACKNOWLEDGMENTSThe authors are grateful to the financial aid from the NFSC(Nos. 21622104, 21871141, 21871142, 21576049, and21701085), the NSF of Jiangsu Province of China (Nos.SBK2017040708 and BK20171032), the Natural ScienceResearch of Jiangsu Higher Education Institutions of China(No. 17KJB150025), Priority Academic Program Develop-ment of Jiangsu Higher Education Institutions, and theFoundation of Jiangsu Collaborative Innovation Center ofBiomedical Functional Materials.

■ REFERENCES(1) Steinlechner, C.; Junge, H. Renewable Methane Generation fromCarbon Dioxide and Sunlight. Angew. Chem., Int. Ed. 2017, 57, 44−45.(2) Inoue, T.; Fujishima, A.; Konishi, S.; Honda, K. Photo-electrocatalytic Reduction of Carbon Dioxide in Aqueous Suspensionsof Semiconductor Powders. Nature 1979, 277, 637−638.(3) Mikkelsen, M.; Jørgensen, M.; Krebs, F. C. The teratonchallenge. A Review of Fixation and Transformation of CarbonDioxide. Energy Environ. Sci. 2010, 3, 43−81.(4) Quadrelli, E. A.; Centi, G. Green Carbon Dioxide. ChemSusChem2011, 4, 1179−1181.(5) Rao, H.; Schmidt, L. C.; Bonin, J.; Robert, M. Visible-Light-Driven Methane Formation from CO2 with a Molecular Iron Catalyst.Nature 2017, 548, 74−77.(6) Stone, E. J.; Lowe, J. A.; Shine, K. P. The Impact of CarbonCapture and Storage on Climate. Energy Environ. Sci. 2009, 2, 81−91.(7) Hull, J. F.; Himeda, Y.; Wang, W.-H.; Hashiguchi, B.; Periana,R.; Szalda, D. J.; Muckerman, J. T.; Fujita, E. Reversible HydrogenStorage Using CO2 and a Proton-switchable Iridium Catalyst inAqueous Media under Mild Temperatures and Pressures. Nat. Chem.2012, 4, 383−388.(8) Tamaki, Y.; Watanabe, K.; Koike, K.; Inoue, H.; Morimoto, T.;Ishitani, O. Development of Highly Efficient Supramolecular CO2Reduction Photocatalysts with High Turnover Frequency andDurability. Faraday Discuss. 2012, 155, 115−127.(9) Liu, Q.; Cong, H. J.; Deng, H. X. Deciphering the SpatialArrangement of Metals and Correla-tion to Reactivity in MultivariateMetal-Organic Frameworks. J. Am. Chem. Soc. 2016, 138, 13822−13825.

Figure 4. Proposed mechanism for the photocatalytic reduction ofCO2 to HCOOH photocatalyzed by NNU-29.

ACS Applied Materials & Interfaces Research Article

DOI: 10.1021/acsami.9b03861ACS Appl. Mater. Interfaces 2019, 11, 25790−25795

25794

(10) Wang, X.; Wisser, F. M.; Canivet, J.; Fontecave, M.; Mellot-Draznieks, C. Immobilization of a Full Photosystem in the Large PoreMIL101 Metal-Organic Framework for CO2 Reduction. ChemSu-sChem 2018, 11, 3315−3322.(11) Liao, P. Q.; Huang, N. Y.; Zhang, W. X.; Zhang, J. P.; Chen, X.M. Controlling Guest Conformation for Efficient Purification ofButadiene. Science 2017, 356, 1193−1196.(12) Zhang, H.; Nai, J.; Yu, L.; Lou, X. W. Metal-Organic-Framework-Based Materials as Platforms for Renewable Energy andEnvironmental Applications. Joule 2017, 1, 77−107.(13) Dhakshinamoorthy, A.; Asiri, A. M.; García, H. Metal−OrganicFramework (MOF) Compounds: Photocatalysts for Redox Reactionsand Solar Fuel Production. Angew. Chem., Int. Ed. 2016, 55, 5414−5445.(14) Li, N.; Liu, J.; Liu, J.-J.; Dong, L.-Z.; Xin, Z.-F.; Teng, Y.-L.;Lan, Y.-Q. Adenine Components in Biomimetic Metal−OrganicFrameworks for Efficient CO2 Photoconversion. Angew. Chem., Int.Ed. 2019, 58, 5226−5231.(15) Xu, H. Q.; Hu, J.; Wang, D.; Li, Z.; Zhang, Q.; Luo, Y.; Yu, S.H.; Jiang, H. L. Visible-Light Photoreduction of CO2 in a Metal−Organic Framework: Boosting Electron−Hole Separation via ElectronTrap States. J. Am. Chem. Soc. 2015, 137, 13440−13443.(16) Luo, T.; Zhang, J.; Li, W.; He, Z.; Sun, X.; Shi, J.; Shao, D.;Zhang, B.; Tan, X.; Han, B. Metal-Organic Framework StabilizedCO2/Water Interfacial Route for Photocatalytic CO2 Conversion.ACS Appl. Mater. Interfaces 2017, 9, 41594−41598.(17) Moghadam, P. Z.; Fairenjimenez, D.; Snurr, R. Q. EfficientIdentification of Hydrophobic MOFs: Application in the Capture ofToxic Industrial Chemicals. J. Mater. Chem. A 2016, 4, 529−536.(18) Ding, N.; Li, H.; Feng, X.; Wang, Q.; Wang, S.; Ma, L.; Zhou,J.; Wang, B. Partitioning MOF-5 into Confined and HydrophobicCompartments for Carbon Capture under Humid Conditions. J. Am.Chem. Soc. 2016, 138, 10100−10103.(19) Mukherjee, S.; Kansara, A. M.; Saha, D.; Gonnade, R.;Mullangi, D.; Manna, B.; Desai, A. V.; Thorat, S. H.; Singh, P. S.;Mukherjee, A.; Ghosh, S. K. An Ultrahydrophobic Fluorous Metal−Organic Framework Derived Recyclable Composite as a PromisingPlatform to Tackle Marine Oil Spills. Chemistry 2016, 22, 10937−10943.(20) Rana, S.; Rajendra, R.; Dhara, B.; Jha, P. K.; Ballav, N. HighlyHydrophobic and Chemically Rectifiable Surface-Anchored Metal-Organic Framework Thin-Film Devices. Adv. Mater. Interfaces 2016,3, No. 1500738.(21) Hu, J.; Ji, Y.; Chen, W.; Streb, C.; Song, Y. “Wiring” redox-active Polyoxometalates to Carbon Nanotubes using a Sonication-driven Periodic Functionalization Strategy. Energy Environ. Sci. 2016,9, 1095−1101.(22) Li, T.; Wang, Z.; Chen, W.; Miras, H. N.; Song, Y. F. RationalDesign of a Polyoxometalate Intercalated Layered Double Hydroxide:Highly Efficient Catalytic Epoxidation of Allylic Alcohols under Mildand Solvent-Free Conditions. Chem. Eur. J. 2017, 23, 1069−1077.(23) Song, Y. F.; Tsunashima, R. Recent Advances on Poly-oxometalate-based Molecular and Composite Materials. Chem. Soc.Rev. 2012, 41, 7384−7402.(24) Ji, Y.; Huang, L.; Hu, J.; Streb, C.; Song, Y. F. Polyoxometalate-functionalized Nanocarbon Materials for Energy Conversion, EnergyStorage and Sensor Systems. Energy Environ. Sci. 2015, 8, 776−789.(25) Du, D. Y.; Qin, J. S.; Li, S. L.; Su, J. M.; Lan, Y. Q. RecentAdvances in Porous Polyoxometalate-Based Metal-Organic Frame-work Materials. Chem. Soc. Rev. 2014, 43, 4615−4632.(26) Xie, S. L.; Liu, J.; Dong, L. Z.; Li, S. L.; Lan, Y. Q.; Su, Z. M.Hetero-Metallic Active Sites Coupled with Strongly ReductivePolyoxometalate for Selective Photocatalytic CO2-to-CH4 Conversionin Water. Chem. Sci. 2018, 10, 185−190.(27) Rousseau, G.; Rodriguez-Albelo, L. M.; Salomon, W.; Mialane,P.; Marrot, J.; Doungmene, F.; Mbomekalle, I. l.-M.; de Oliveira, P.;Dolbecq, A. Tuning the Dimensionality of Polyoxometalate-BasedMaterials by Using a Mixture of Ligands. Cryst. Growth Des. 2015, 15,449−456.

(28) Rodriguez-Albelo, L. M.; Ruiz-Salvador, A. R.; Sampieri, A.;Lewis, D. W.; Gomez, A.; Nohra, B.; Mialane, P.; Marrot, J.;Secheresse, F.; Mellot-Draznieks, C.; Ngo Biboum, R.; Keita, B.;Nadjo, L.; Dolbecq, A. Zeolitic Polyoxometalate-Based Metal-OrganicFrameworks (Z-POMOFs): Computational Evaluation of Hypo-thetical Polymorphs and the Successful Targeted Synthesis of theRedox-Active Z-POMOF1. J. Am. Chem. Soc. 2009, 131, 16078−16087.(29) Nohra, B.; El Moll, H.; Rodriguez Albelo, L. M.; Mialane, P.;Marrot, J.; Mellot-Draznieks, C.; O’Keeffe, M.; Ngo Biboum, R.;Lemaire, J.; Keita, B.; Nadjo, L.; Dolbecq, A. Polyoxometalate-BasedMetal Organic Frameworks (POMOFs): Structural Trends, Ener-getics, and High Electrocatalytic Efficiency for Hydrogen EvolutionReaction. J. Am. Chem. Soc. 2011, 133, 13363−13374.(30) Rodriguez Albelo, L. M.; Ruiz-Salvador, A. R.; Lewis, D. W.;Gomez, A.; Mialane, P.; Marrot, J.; Dolbecq, A.; Sampieri, A.; Mellot-Draznieks, C. Zeolitic Polyoxometalates Metal Organic Frameworks(Z-POMOF) with Imidazole Ligands and ε-Keggin Ions as BuildingBlocks; Computational Evaluation of Hypothetical Polymorphs and aSynthesis Approach. Phys. Chem. Chem. Phys. 2010, 12, 8632−8639.(31) Qin, J. S.; Du, D. Y.; Guan, W.; Bo, X. J.; Li, Y. F.; Guo, L. P.;Su, Z. M.; Wang, Y. Y.; Lan, Y. Q.; Zhou, H. C. UltrastablePolymolybdate-Based Metal−Organic Frameworks as Highly ActiveElectrocatalysts for Hydrogen Generation from Water. J. Am. Chem.Soc. 2015, 137, 7169−7177.(32) Tamaki, Y.; Morimoto, T.; Koike, K.; Ishitani, O. Photo-catalytic CO2 Reduction with High Turnover Frequency andSelectivity of Formic Acid Formation Using Ru(II) MultinuclearComplexes. Proc. Natl. Acad. Sci. U.S.A. 2012, 109, 15673−15678.(33) Lefebvre, J.-F.; Schindler, J.; Traber, P.; Zhang, Y.; Kupfer, S.;Gra fe, S.; Baussanne, I.; Demeunynck, M.; Mouesca, J.-M.;Gambarelli, S.; Artero, V.; Dietzek, B.; Chavarot-Kerlidou, M. Anartificial photosynthetic system for photoaccumulation of twoelectrons on a fused dipyridophenazine (dppz)−pyridoquinolinoneligand. Chem. Sci. 2018, 9, 4152−4159.(34) Wang, S.; Yao, W.; Lin, J.; Ding, Z.; Wang, X. CobaltImidazolate Metal-Organic Frameworks Photosplit CO2 under MildReaction Conditions. Angew. Chem., Int. Ed. 2014, 53, 1034−1038.(35) Pan, F.; Li, B.; Xiang, X.; Wang, G.; Li, Y. Efficient CO2Electroreduction by Highly Dense and Active Pyridinic Nitrogen onHoley Carbon Layers with Fluorine Engineering. ACS Catal. 2019, 9,2124−2133.(36) Wang, Y.; Chen, J.; Wang, G.; Li, Y.; Wen, Z. PerfluorinatedCovalent Triazine Framework Derived Hybrids for the HighlySelective Electroconversion of Carbon Dioxide into Methane.Angew. Chem., Int. Ed. 2018, 57, 13120−13124.

ACS Applied Materials & Interfaces Research Article

DOI: 10.1021/acsami.9b03861ACS Appl. Mater. Interfaces 2019, 11, 25790−25795

25795